U.S. patent application number 12/134847 was filed with the patent office on 2009-05-14 for method for determining temporal solar irradiance values.
Invention is credited to Aaron William Woro.
Application Number | 20090125275 12/134847 |
Document ID | / |
Family ID | 41398413 |
Filed Date | 2009-05-14 |
United States Patent
Application |
20090125275 |
Kind Code |
A1 |
Woro; Aaron William |
May 14, 2009 |
Method For Determining Temporal Solar Irradiance Values
Abstract
A method for generating temporal solar irradiance values for a
selected area. Binary format hillshade files are generated for
selected azimuth and altitude points on the Sun's path for selected
time points for the area. Data in the hillshade files is
reclassified into reclassified files, on basis of the selected time
points relative to the solar radiation data. The reclassified files
are then summed to generate a set of normalized reclassified files,
each representing a selected intermediate interval. The values for
each corresponding one of the cells in the set of normalized
reclassified files are summed to generate an irradiance-weighted
shade file. The hillshade files are summed by inclusively OR-ing
corresponding values for each of the cells in each of the hillshade
files to generate respective composite files for each said selected
intermediate interval. The composite files are then summed to
generate a summed shade/time frequency file in which each data
point therein represents the frequency of repetition of
corresponding cells in the hillshade files over a selected upper
interval of time. Each data point value in the irradiance-weighted
shade file is then divided by the corresponding data point value in
the frequency file to generate a file having solar access values
for the upper interval, relative to the intermediate interval, for
the selected area.
Inventors: |
Woro; Aaron William;
(Boulder, CO) |
Correspondence
Address: |
LATHROP & GAGE LC
2345 GRAND AVENUE, SUITE 2800
KANSAS CITY
MO
64108
US
|
Family ID: |
41398413 |
Appl. No.: |
12/134847 |
Filed: |
June 6, 2008 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11740087 |
Apr 25, 2007 |
7500391 |
|
|
12134847 |
|
|
|
|
Current U.S.
Class: |
702/182 ;
73/170.27 |
Current CPC
Class: |
G06F 2111/10 20200101;
Y02B 10/20 20130101; G06F 30/13 20200101; F24S 40/20 20180501; G06F
2119/06 20200101; G06F 30/20 20200101; F24S 2020/16 20180501; F24S
2201/00 20180501 |
Class at
Publication: |
702/182 ;
73/170.27 |
International
Class: |
G21C 17/00 20060101
G21C017/00; F24J 2/00 20060101 F24J002/00 |
Claims
1. A method for generating temporal solar irradiance values
comprising: generating binary format hillshade files, for selected
azimuth and altitude points on the Sun's path for selected time
points, wherein each of the hillshade files comprises a matrix of
data cells representing shaded relief patterns for a selected area;
reclassifying data in the hillshade files into reclassified files,
on basis of the selected time points relative to the solar
radiation data, wherein each of the hillshade files includes data
for one of a selected number of different time points for a
selected number of intermediate intervals in a selected upper
interval; summing the reclassified files to generate a set of
normalized reclassified files; summing the values for each
corresponding one of the cells in the set of normalized
reclassified files to generate an irradiance-weighted shade file;
summing the hillshade files, by inclusively OR-ing corresponding
values for each of the cells in each of the hillshade files to
generate respective composite files for each of the selected
intermediate intervals; summing the composite files to generate a
summed shade/time frequency file in which each data point therein
represents the frequency of repetition of corresponding cells in
the hillshade files over the selected upper interval of time; and
dividing each data point value in the irradiance-weighted shade
file by the corresponding data point value in the frequency file to
generate a file comprising solar access values for the upper
interval, relative to the intermediate interval, for the selected
area.
2. The method of claim 1, wherein each of the cells in each of the
hillshade files contains a binary number representing either a
shadow or a non-shadow condition of a specific X/Y coordinate
within the selected area for a specific time.
3. The method of claim 1, wherein each of the hillshade files
includes data for the same time points for different ones of the
intermediate intervals in the upper interval.
4. A method for generating temporal solar irradiance values
comprising: generating binary format hillshade files, for selected
azimuth and altitude points on the Sun's path for selected time
points comprising, wherein each of the hillshade files comprises a
matrix of data cells representing shaded relief patterns for a
selected area; reclassifying data in the hillshade files into
reclassified files, on basis of the selected time points relative
to the solar radiation data; summing the reclassified files to
generate a set of normalized reclassified files each representing a
selected intermediate interval; summing the values for each
corresponding one of the cells in the set of normalized
reclassified files to generate an irradiance-weighted shade file;
summing the hillshade files, by inclusively OR-ing corresponding
values for each of the cells in each of the hillshade files to
generate respective composite files for each said selected
intermediate interval; summing the composite files to generate a
summed shade/time frequency file in which each data point therein
represents the frequency of repetition of corresponding cells in
the hillshade files over a selected upper interval of time; and
dividing each data point value in the irradiance-weighted shade
file by the corresponding data point value in the frequency file to
generate a file comprising solar access values for the upper
interval, relative to the intermediate interval, for the selected
area.
5. The method of claim 4, wherein each of the cells in each of the
hillshade files contains a binary number representing either a
shadow or a non-shadow condition of a specific X/Y coordinate
within the selected area for a specific time.
6. The method of claim 4, wherein each of the hillshade files
includes data for the same time points for different ones of the
intermediate intervals in the upper interval.
7. The method of claim 4, wherein each of the hillshade files
includes data for one of a selected number of different time points
for a selected number of intermediate intervals in a selected upper
interval.
8. A method for generating temporal solar irradiance values for a
selected area comprising: generating a plurality of hillshade
files, using X/Y/Z digital elevation model data, for a plurality of
points representing the Sun's azimuth and altitude relative to the
selected area, using values for astronomical positions of the sun
relative to the latitude and longitude for each of the points,
wherein each of the hillshade files comprises a matrix of cells
containing binary numbers, each representing either a shadow or a
non-shadow condition of a specific X/Y coordinate within the
selected area for a specific time; reclassifying each of the
hillshade files, based on irradiance levels, to generate
reclassified hillshade files, wherein each of the cells therein
represents the percentage of total radiation during a respective
one of the lower intervals for the selected area; generating a set
of normalized reclassified files by summing groups of the
reclassified hillshade files, wherein each file in the set
represents the sum of all of the associated normalized files within
at least one selected intermediate interval of time; summing the
values for each corresponding one of the cells in the set of
normalized reclassified files, within a selected upper interval of
time represented by the set of normalized reclassified files, to
generate an irradiance-weighted shade file for the upper interval;
summing each of the hillshade files, by inclusively OR-ing
corresponding values for each of the cells in each of the hillshade
files to generate a set of composite files for the selected
intermediate intervals; summing set of composite files to generate
a single composite hillshade frequency file in which each data
point therein represents the frequency of repetition, of
corresponding said cells in the hillshade files for the lower
interval, over the upper interval; and dividing the
irradiance-weighted shade file by the summed shade/time frequency
file, using a matrix division operation, to provide a file
indicating the average percent of solar access per said cell for
the upper interval, relative to the intermediate interval.
9. The method of claim 8, wherein each of the hillshade files
includes data for the same time points for different ones of the
intermediate intervals in the upper interval.
10. The method of claim 8, wherein each of the hillshade files
includes data for one of a selected number of different time points
for a selected number of intermediate intervals in a selected upper
interval.
11. A method for predicting the total output of a solar energy
system over a selected interval of time comprising: generating
binary format hillshade files, for selected azimuth and altitude
points on the Sun's path, for selected times within the interval,
wherein each of the hillshade files comprises a matrix of data
cells representing shaded relief patterns for an area covered by
the solar energy system; reclassifying data in the hillshade files
into reclassified files, on basis of selected time points relative
to the solar radiation data; summing the reclassified files to
generate a set of normalized reclassified files; and calculating
time-dependent irradiance values for the area covered by the solar
energy system by summing the values for each corresponding one of
the cells in the set of normalized reclassified files to generate
an irradiance-weighted shade file indicating the total output of
the solar energy system over the selected interval of time.
12. The method of claim 11, wherein each of the cells in each of
the hillshade files contains a binary number representing either a
shadow or a non-shadow condition of a specific X/Y coordinate
within the selected area for a specific time.
13. A method for generating temporal solar irradiance values
comprising: collecting three-dimensional aerial data of a selected
area; generating a digital elevation model of the selected area
from the aerial data; generating obstruction shadows by creating a
hillshade file for each obstruction casting a shadow onto a portion
of the selected area, using solar positions calculated at intervals
over a specific period of time; determining shadow-free areas, in
the selected area, from the digital elevation model and the
obstruction shadows by intersecting the usable area remains with
total roof area, for each of the rooftops, to generate shadow-free
areas on the rooftops that are free of said obstruction shadows
over the specific period; wherein the shadow-free areas are
determined by: generating a virtual city for the selected area
using data from a digital elevation model, integrated with data
from real-world features from the aerial data, of the selected
area; extracting the rooftop shapes as three-dimensional data, from
the virtual city; and calculating the shade cast by obstructions
from the shadow simulation data, by: reclassifying data in the
hillshade files into reclassified files, on basis of the selected
time points relative to the solar radiation data; summing the
reclassified files to generate a set of normalized reclassified
files each representing a selected intermediate interval; summing
the values for each corresponding one of the cells in the set of
normalized reclassified files to generate an irradiance-weighted
shade file; summing the hillshade files, by inclusively OR-ing
corresponding values for each of the cells in each of the hillshade
files to generate respective composite files for each said selected
intermediate interval; summing the composite files to generate a
summed shade/time frequency file in which each data point therein
represents the frequency of repetition of corresponding cells in
the hillshade files over a selected upper interval of time; and
dividing each data point value in the irradiance-weighted shade
file by the corresponding data point value in the frequency file to
generate a file comprising the shade cast by obstructions for the
upper interval, relative to the intermediate interval, for the
selected area.
14. The method of claim 13, wherein each of the cells in each of
the hillshade files contains a binary number representing either a
shadow or a non-shadow condition of a specific X/Y coordinate
within the selected area for a specific time.
15. The method of claim 13, wherein each of the hillshade files
includes data for the same time points for different ones of the
intermediate intervals in the upper interval.
16. The method of claim 13, wherein each of the hillshade files
includes data for one of a selected number of different time points
for a selected number of intermediate intervals in a selected upper
interval.
Description
RELATED APPLICATIONS
[0001] The present application is a continuation-in-part of U.S.
patent application Ser. No. 11/740,087, filed Apr. 25, 2007, titled
"System and Method for Identifying the Solar Potential of
Rooftops".
BACKGROUND
[0002] The evaluation method for profiling a house's potential for
solar power is presently manual labor-intensive. A typical house
evaluation includes the required instruments and a climb atop the
roof to profile the solar power potential by estimated k/W hours
and expected return on investment following the assessment of roof
geometry and the effect on shading from obstructions. If the house
has the necessary geometric specifications and reasonable annual
solar exposure, then a certain solar power system size is
recommended. However, the process is time-consuming and relatively
expensive.
[0003] A typical homeowner who is considering solar power must
first find a listing of solar providers, and then arrange a time
and date for each one of them to evaluate the house for potential.
This often occurs as many as five separate times, as opinions and
measurements differ between evaluations. The aggregate effect on
the solar industry is considerable. From the initial inquiry
through to installation, the resources spent on evaluations can run
into the hundreds of dollars. This cost is ultimately passed on to
consumers, for whom this is often the chief obstacle when
considering solar power. In other cases, those who own houses that
are perfectly situated to exploit the benefits of distributed solar
power are often unaware of the investment potential. Currently,
targeted demographic marketing is the primary tool used in the
expansion of solar power.
[0004] The present system and method solves a key problem in
residential solar power distribution by removing the need for
physical site visits to survey and qualify houses for potential
solar power utilization. Using various types of satellite and
remote-sensing technology in addition to datasets already made
available by the United States Census Bureau, commercial
demographic services, and municipal parcel databases, the present
system employs GIS or IGS (Interactive Geometric Software) software
to analyze numerous variables, in three dimensions, related to a
given dwelling, including the surface area of its roof; the slope
and direction of the roof's various portions; related address and
owner information for each dwelling; also, items specific to the
microclimate and precise geographic location of the structure are
taken into account by the process including potential shading from
trees and other local structures and obstructions. The information
required as input data for the GIS system is obtained using aerial
remote sensing technology.
SUMMARY
[0005] A method is disclosed for generating temporal solar
irradiance values. Initially, binary format hillshade files are
generated for selected azimuth and altitude points on the Sun's
path for selected time points for a selected area. Data in the
hillshade files is reclassified into reclassified files, on basis
of the selected time points relative to the solar radiation data.
The reclassified files are then summed to generate a set of
normalized reclassified files, each representing a selected
intermediate interval. The values for each corresponding one of the
cells in the set of normalized reclassified files are summed to
generate an irradiance-weighted shade file. The hillshade files are
summed by inclusively OR-ing corresponding values for each of the
cells in each of the hillshade files to generate respective
composite files for each said selected intermediate interval. The
composite files are then summed to generate a summed shade/time
frequency file in which each data point therein represents the
frequency of repetition of corresponding cells in the hillshade
files over a selected upper interval of time. Each data point value
in the irradiance-weighted shade file is then divided by the
corresponding data point value in the frequency file to generate a
file having solar access values for the upper interval, relative to
the intermediate interval, for the selected area.
BRIEF DESCRIPTION OF THE DRAWINGS
[0006] FIG. 1 is a flowchart showing a high-level exemplary
embodiment of the present system;
[0007] FIG. 2A is a flowchart showing a mid-level exemplary
embodiment of the present system;
[0008] FIG. 2B is a diagram showing a digital elevation/terrain
model in terms of XYZ coordinates;
[0009] FIG. 3 is a flowchart showing an exemplary embodiment of the
present system in greater detail;
[0010] FIG. 4 is a flowchart showing exemplary details of process
112 for extracting buildings and rooftops as three-dimensional
vector or raster files;
[0011] FIG. 5 is a flowchart showing exemplary datasets that serve
as inputs to the present system;
[0012] FIG. 6 is a flowchart showing exemplary details of process
114 for roof portion-specific slope extraction;
[0013] FIG. 7 is a diagram showing an example of process 116 for
determining roof orientation;
[0014] FIG. 8 is a flowchart showing an exemplary method for
determining annual tree shade patterns;
[0015] FIG. 9 is a flowchart showing exemplary details of the
process for converting annual shadow extraction to usable roof
area;
[0016] FIG. 10 is a flowchart showing exemplary details of a
process for generating cartographic data;
[0017] FIG. 11 is a flowchart showing exemplary details of a
process for determining solar access values for specified intervals
of time;
[0018] FIG. 12 is a diagram showing examples of data formats
corresponding to the process shown in FIG. 11;
[0019] FIG. 13 is an exemplary diagram showing the extraction of a
Digital Elevation Model from raw XYZ data;
[0020] FIG. 14A is an exemplary diagram showing an example of a
hillshade file for a specific area;
[0021] FIG. 14B shows the diagram of FIG. 14A in a binary
format;
[0022] FIG. 15 is an exemplary diagram showing the raw binary data
for a hillshade file superimposed over the original DEM data for
the same area;
[0023] FIG. 16 is an exemplary diagram showing hillshade file
binary data for the area over rooftop 1310;
[0024] FIGS. 17A, 17B, and 17C are exemplary diagrams showing three
respective sets of reclassified data for the area over rooftop
1310;
[0025] FIG. 18 is an exemplary diagram showing summed data 1801
from sets 1701, 1702, and 1703;
[0026] FIG. 19A is an exemplary diagram showing the generation of a
composite file from a file set which includes sets of hillshade
files;
[0027] FIG. 19B is an exemplary diagram showing the process of
composite-to-new raster generation;
[0028] FIG. 20 is an exemplary diagram showing summed shade/time
frequency file 2001 generated from file sets 1701, 1702, and 1703,
representing the denominator of equation 1130; and
[0029] FIG. 21 is an exemplary diagram showing resultant data 2101
after division operation 1131 has been performed.
DETAILED DESCRIPTION
[0030] FIG. 1 is a flowchart showing a high-level exemplary
embodiment of the present system and method. The present GIS-based
technology and associated algorithms deliver an output of specific
addresses in both cartographic and tabular form that reveal the
houses (or other buildings) found to match the defined parameters
set by the user. The information delivered also contains the
specific data for solar potential for every dwelling within the
coverage.
[0031] As shown in FIG. 1, at step 101, public and commercial
spatial and tabular data are entered into a geographic information
system (`GIS`). The present system analyzes the solar potential of
houses and other buildings based on variable parameters including
some or all of the following specifications and other data:
[0032] Roof space
[0033] Roof direction
[0034] Roof slope
[0035] Address and demographic profiles
[0036] Lack of tree cover (annual shadow patterns over
rooftops)
[0037] Listed below are examples of public/commercial datasets that
may be used as system inputs in the solar potential evaluation
process:
[0038] Parcel frameworks (i.e., property boundaries)
[0039] Digital Elevation Models (used for forestry, flood plain
mapping, etc.)
[0040] High-resolution (<30 cm) aerial photography
[0041] Climate inclusive solar irradiance levels
[0042] At step 102, aerial data is collected. This data may be
efficiently collected by an airplane, but may, alternatively, be
collected by other types of aircraft (e.g., balloons) or
Earth-orbiting vehicles (spacecraft). In an exemplary embodiment,
aerial data includes LiDar (Laser Detection and Ranging), LiDar
infrared imaging, and aerial photography of parcels of land on
which houses and other buildings are located. A digital elevation
model (DEM) and/or a digital terrain model (DTM) are extracted from
the LiDar and aerial photographs, which are often acquired
concurrently during the same flight missions.
[0043] An established/standardized GIS-based algorithm set is run
to receive data and to allow for geoprocessing, and delivers
addresses and precise profiles of houses in a given county (or
other areas) that demonstrate various levels of potential for
distributed solar power in their entirety, or as specified by the
user and/or commercial/government entity making the inquiry. At
step 103, algorithm sets (described in detail below) are then run
on the aerial and GIS data to generate data 135 in cartographic and
tabular form of the solar potential of houses, with specifications
by `parcel` address, i.e., specifications for a particular plot of
land on which a dwelling is located.
[0044] The GIS algorithms receive and process the necessary data,
which is consistently similar in data requirements that are entered
into a geodatabase. The data is then processed (georeferenced,
georectified, tables constructed, layering performed, etc.) and
processed by the present system.
[0045] FIG. 2A is a flowchart showing a mid-level exemplary
embodiment of the present system. The data collection method for
the present geospatial analysis employs remote sensing technology
to extract real-world features in three dimensions, which are
stored as X, Y, and Z positions of a spatial coverage (or using an
equivalent spatial storage method). As shown in FIG. 2, in block
105, aerial data is collected via aircraft (and/or spacecraft).
More specifically, in an exemplary embodiment, the information
required as input data for the present GIS system is obtained using
airborne remote sensing technology, such as LiDar (Light Detection
and Ranging), IFSAR/INSAR (Interferometric Synthetic Aperture
Radar), infrared spectrometry, aerial photogrammetry, etc. This
input data includes real-world features 106 such as vegetation
canopy, buildings, and clutter.
[0046] Raw data 107 from digital elevation/terrain models (DEM/DTM)
142, in conjunction with data from real-world features 106, is then
used to generate a `virtual city` 111, at step 109 A virtual city
is a digital depiction of real world geography and features,
comprising a three-dimensional dataset extracted from one or more
digital elevation models via photogrammetry. The digital
elevation/terrain models are extracted using LiDar/IFSAR, and may
also be generated from other passive/active forms of high
resolution remote sensing. Aerial/photogrammetric DEM extraction
may be used to generate the digital elevation/terrain models if the
technology employed is sufficiently accurate.
[0047] FIG. 2B is a diagram of two elements of virtual city 111
showing example XYZ coordinates (which are employed in all digital
elevation/terrain models), using a single building 201 and tree 202
as an example of data included in virtual city 111, comprising
DEM/DTM data 142 that has been integrated with data from real-world
features 106. As shown in FIG. 2B, the top center 201(T) of
building 201 is represented by coordinates X1, Y1, Z1, and the
highest point 202(T) of tree 202 is represented by coordinates X2,
Y2, Z2.
[0048] The furthest point C of shadow 203 from high point 202(T) of
tree 202 is used in determining annual shading effects. For any
given date/time, point C can be determined with reference to the
triangle ABC, where the angle ACB of the triangle (relative to the
horizontal) is determined by the line passing between the sun 205,
point 202(T), and the rooftop of building 201. In the example of
FIG. 2B, given angle ACB, the length 204 of hypotenuse BC of
triangle ABC can readily be determined from the height of the roof,
the height of point 202(T), and the horizontal distance AC between
point 202(T) and point C. The potential tree shadow arc with
respect to the rooftop of building 201 can be built on multiple
separate calculations (one for each desired interval), as described
below in detail with respect to FIGS. 8 and 9.
[0049] As shown in FIG. 2A, virtual city DEM/DTM data 111 is
processed by GIS system 110 (shown in FIG. 3, described below) to
generate tabular and cartographic data 133. More specifically, in
step 115, building, roof, and rooftop shape extraction is
performed. This process is described in detail below with respect
to FIGS. 6, 7, and 8.
[0050] In step 120, solar azimuth and altitude angles are
calculated for desired intervals (e.g., every hour over an
annual/monthly/weekly time period) to determine the shadow 122 cast
by obstructions including vegetation, buildings and other
structures. The usable remains 125 of the rooftop sections are then
generated, and then the XY shadow simulation data is intersected
with the XY coordinates of the rooftop areas, in step 118, to yield
overlaid files, having spatial attributes, showing rooftop areas
that are shade-free (or which have shadow cover less than a
pre-specified percentage of time) over the course of a year or
other specific period of time. All points/pixels that exist outside
these coordinate boundaries are deleted. This process is described
in detail below with respect to FIG. 9.
[0051] Next, the outputs of process steps 115 and 118 are formatted
in both tabular and cartographic form, at step 133. The tabular and
cartographic data is then filtered as requested by a system user in
step 134 to provide solar census output 135.
[0052] FIG. 3 is a flowchart showing an exemplary embodiment of the
present system in greater detail. As shown in FIG. 3, data
collection is performed by aircraft 160 (and/or spacecraft) at step
105, as described above with respect to FIG. 2A. Raw data 107 from
digital elevation/terrain models (DEM/DTM) 142, in conjunction with
real-world features 106, is then used to generate `virtual city`
111, the data for which is transmitted from the aircraft 160 to GIS
110.
[0053] Extracting buildings/rooftops as three-dimensional vector
files is the first preferred step of geographic/geometric
deduction, wherein rooftops are identified by their location and
the rest of the features and corresponding XY coordinates are
filtered out of the geospatial data. FIG. 4 is a flowchart showing
exemplary details of process 112 (in FIG. 3) for extracting
buildings and rooftops as three-dimensional vector files. Operation
of the present system is best understood by viewing FIGS. 3 and 4
in conjunction with one another. As shown in FIG. 4, at step 405,
DEM analysis software is configured to the GIS 100. Examples of DEM
analysis software are LiDar Analyst.TM. by Visual Learning Systems,
and SilverEye.TM. by GeoTango International Corp. This software
enables the extraction of building/spatial features using
automated/semi-automated processes.
[0054] In step 410, the Digital Elevation Model file is loaded into
GIS system 110 for analysis. Z (height) data values must be
available to provide attribute differentials during extraction. In
step 415, the identification of building features is ordered from
the XYZ model (DEM/DTM) using the necessary spatial parameters.
Spatial parameters are necessary to enable geometric processing. In
the case of residential rooftops, these spatial parameters may
include, for example: 1) size of building surface->10
m.sup.2<40 m.sup.2; 2) slope
parameters->5.degree.-<40.degree.; 3) texture variance, which
assists in differentiating between trees, clutter, buildings etc.;
and 4) distance above recognized surface (e.g., distance above
ground)->3 m-<15 m. When analyzing the solar potential of
commercial rooftops, these parameters may vary (e.g., slope is
often 0.degree., area is >100 m.sup.2, etc.).
[0055] In step 420, rooftop polygons are extracted as
multi-component sections, in order to allow for further processing.
In step 425, the polygons (which, in Geographic Information
Systems, are files in vector format typically referred to as
`shapefiles`) are saved with attributes, which may include
calculated areas for multiple roof components, roof type (e.g.,
gabled, complex etc.), general polygon slope, and which must
include X,Y, and Z values. Typical attributes may include roof
type, slope, area, etc. In step 430, these shapefiles are imported
into geodatabase 104, with multi-component rooftop FIDs (feature
IDs) joined to parcel IDs and georeferenced to all other data.
[0056] Each of the shapefiles contains attributes, but none are
bound to the output without further processing, which is necessary
to produce data in geospatial/cartographic format for use in the
processes that follow. Shown in FIG. 5 is the concept 500 of
datasets that serve as inputs to the present system 100. In an
exemplary embodiment, these datasets may include parcel-specific
LiDar or other DEM (XYZ) data 513, parcel building frame data,
parcel-specific census data, and parcel-specific electrical
usage.
[0057] The control arrow between blocks 513 and 522 signifies the
processing that enables these datasets to serve as bound
cartographic frameworks, offers precise readings of roof
specifications and shade, and allow for restrictions to be placed
on the large demographic datasets that will reveal only the parcels
that adhere to the geodemographic parameters set and/or input by a
system user. Parcel building frame data 511 and other spatial
requirements 512 are processed to generate cartographic frameworks
521, and, in one embodiment, parcel-specific LiDar data 513 is used
to generate geometric display data 522 of roof specifications and
tree cover. Parcel-specific census data and parcel-specific
electrical usage are used to generate demographic restrictions 523,
such as parcel value, home ownership, household income etc.
[0058] The processed data described above is registered in the
geodatabase 104 as specific addresses that have satisfied criteria
used by processing algorithms to reveal the parcels (home sites)
that meet a threshold solar potential. Cartographic framework 521
and display data 522 are combined to generate parcel-specific
cartographic data 132 (FIGS. 3 and 5), and demographic data and
restrictions 523 are used to generate parcel-specific text data
130. Graphical data 132 and text data 130 comprise the output-ready
solar census 135, stored in database 104.
[0059] In an exemplary embodiment, when database 104 is
queried/filtered to generate the final output, the database output
takes the form of: [0060] (1) A map 139 indicating the precise
locations of the parcels that have met or exceeded the
solar-potential criteria used in the process; and [0061] (2) A
table 138 containing variable parameters (specifications) indicated
above with respect to FIG. 1, and additionally including addresses
for each parcel.
[0062] FIG. 6 is a flowchart showing exemplary details of process
114 (in FIG. 3) for roof portion-specific slope extraction. As
shown in FIG. 6, in step 605, a new slope dataset is created from
raw XYZ file data using the GIS 110. This conversion is applied to
the entire dataset. The new slope layer is then converted to
integer format. In step 610, the dataset is converted to points,
and point-specific attributes are bound to point coverage. There is
now one datum for every point within the coverage.
[0063] In step 615, the extracted rooftop polygons are added to the
GIS workspace, and the attribute-bound slope point data is
intersected with the XY coordinates of the rooftop polygons. All
points that fall outside the polygons are deleted. There is now a
single dataset that represents one slope datum (e.g., 18.degree.)
per point at the given resolution within the rooftop area only.
[0064] In step 620, statistical divisions and anomalies that divide
the separate portions of the roof space are calculated, and the
divided portions are ordered to represent the multi-part rooftops
and bind attribute sub-data to their respective portions on the XY
plane.
[0065] In step 625, the mean spatial statistics of all data within
each roof portion are generated and all attribute data is
agglomerated across the coverage to this point-in-polygon subset to
create an agglomerated attribute file.
[0066] This attribute file is then imported into the geodatabase
104 at step 630, with all rooftop FIDs joined to parcel IDs and
georeferenced. An output example is the following tabular data
indicating parcel address, roof-section polygon feature
identification number, roof-section slope (determined by extracted
Z statistics), and roof section dimensions:
[0067] 29 Palms Ave, Miami Fla. 33107
[0068] FID1--16.degree. (X meters/Y meters)
[0069] FID2--16.degree. (X meters/Y meters)
[0070] FID3--22.degree. (X meters/Y meters)
[0071] FID4--22.degree. (X meters/Y meters)
[0072] FIG. 7 is a diagram showing an example of process 116 for
determining roof orientation. As shown in FIG. 7, in step 705, an
orientation/aspect dataset is created from the XYZ data file using
GIS 110. This conversion is to be applied to the entire dataset;
new orientation data is then converted to text and integer formats.
In step 710, all integer data is converted to points, and
point-specific attributes are bound to point coverage. There is now
one datum for orientation for every point within the coverage.
[0073] In step 715, the extracted rooftop polygons/area boundary
data are added to the GIS workspace, and in step 720, rooftop data
files and orientation point data are georeferenced so that the data
is aligned on the XY plane. In step 725, all integer data is
converted to polygons with bound attributes relevant to orientation
and feature identification.
[0074] In step 730, orientation ranges (e.g.: I, II, and III
(120-170, 171-220, and 221-280) are queried to create new aspect
polygon files for each queried range. The purpose of this operation
is to divide the orientation of the roof areas into individual
files for a DBMS (database management system) such as SQL. The
present step identifies the orientation of a single roof portion,
as almost all rooftops in their entirety are orientated to
360.degree. of the horizontal plane. Extracting the individual
attributes can be done as single integers without queries as well.
In step 735, the attribute-bound orientation point/polygon data is
intersected with the XY coordinates of the rooftop areas only. All
points that fall outside these coordinate boundaries are deleted.
There are now datasets in any division, determined by the user's
DBMS query preferences, that represents one datum (e.g.,
180.degree./south) per point inside the individual rooftop
portions. In step 740, the mean XY statistics of all orientation ID
data within each roof portion are generated, and all attribute data
is agglomerated across the coverage to point-in-polygon subsets
representing range divisions as preferred. The resulting data is
stored in an agglomerated attribute file.
[0075] In step 745, this attribute file is imported into the
geodatabase 104, with all rooftop FIDs joined to parcel IDs and
georeferenced. An output example of tabular data generated from
this attribute file (at step 750) is shown below:
[0076] 29 Palms Ave, Miami Fla. 33107
[0077] FID1--72.degree. ENE
[0078] FID2--162.degree. SSE
[0079] FID3--252.degree. WSW
[0080] FID4--342.degree. NNW
[0081] The GIS database 104 is controlled by queries, the desired
purpose of which is to locate portions of rooftops with optimum
orientation for solar power. However, the present method ultimately
allows for all rooftops to be split and identified by orientation
regardless of user preference.
[0082] FIG. 8 is a diagram showing an exemplary method for
determining annual tree shade patterns. To accurately simulate
annual shade patterns cast by trees and surrounding obstructions, a
tree cover algorithm creates a series of overlapping vector files
around trees identified by, for example, LiDar canopy files.
[0083] A `zenith angle` is the angle between the local zenith and a
line of sight to an object (i.e., the Sun 205). If point C=the
ground site, point Z=any point directly above point C (e.g., the
zenith), and point S=the Sun, then the Sun's viewing zenith
angle=the angle ZAS. The zenith angle z is thus defined to be the
angle subtended by the sun at the center of the earth and normal
(perpendicular) to the surface of the earth. The zenith angle in
the example of FIG. 8 is the angle subtended by the Sun (point S),
point C, and the zenith Z, or angle SCZ, which is equal to angle
(90.degree.-ACB), the complement of the angle between the horizon
and the sun as viewed from the edge of shadow 503 furthest from
tree 505.
[0084] Simulated monthly (or other interval) shadow extent is then
extracted based on the latitude/longitude (XY) location of
individual building and tree height (determined by Z value in the
data). Simple Pythagorean geometry with automated shadow
identification using grey-scale range is employed in the tree cover
algorithm as follows: [0085] 1) GIS 100 locates point A (bottom of
tree/established surface at relative elevation 0.0, the `bare
Earth` surface) at a point on the vegetation. [0086] 2) Point B
(top of tree) is located via LiDar data, using the Z value for that
point. [0087] 3) Tree height is distance 501 between A and B.
[0088] 4) Point C is located by the GIS by determining the precise
XY location on the surface based on the position of the sun
relative to the tree/obstruction casting the shadow (line SBC).
This line is a result of a simulated projection relative to the
obstruction height B and XY surface location A. [0089] 5) The
hypotenuse 504 of triangle ABC and the height of the tree 501 are
extracted from these calculations, using the Pythagorean theorem,
to yield the distance 502 the triangle can cover. A set of
potential tree shadow arcs 503 can then be built on separate
calculations (one for each month, or any other desired interval)
using a geometric algorithm and solar zenith/altitude and azimuth
data within the DEM, as described in FIG. 9. An alternative process
that extracts the same data with the same geometry deploys `line of
sight` geometry to determine the line SBC referred to above,
without employing trigonometry. [0090] 6) The filled-in arc 503 of
annual shadow extent is then generated, using the method employed
in FIG. 9.
[0091] FIG. 9 is a flowchart showing exemplary details of the
process for extracting annual shadow extraction and converting the
data to usable roof area. This flowchart details steps 120, 122,
125, and 118 in FIG. 3, and employs the geometry shown in FIG. 8.
More specifically, this conversion process is performed as
follows.
[0092] In step 905, XYZ/digital elevation file of geographic
coverage of the area for which data has been acquired (or other
selected area of interest) is loaded into GIS 110. In step 910,
solar azimuth and altitude data are entered into GIS 110 to
simulate dynamic celestial effects on surface shade for desired
chronological intervals. For example, a single entry may have a
solar altitude of 59.2 and solar azimuth of 99.2 to simulate the
shading effects over a three dimensional surface at 10:19 am on
Aug. 3, 2006 at 80.degree. W and 25.degree. N.
[0093] To extract shadows cast by obstructions, the GIS/IGS applies
the Pythagorean theorem. If (x0, y0) and (x1, y1) represent points
in the plane, then the shadow cast will be determined by the
altitude and azimuth angles of the sun relative to those
coordinates. The height (Z) of the points creating the obstruction
(e.g., tree top or tree branch, chimney, etc.) from the surface is
(a1-b1).sup.2+the distance from that coordinate on the (XY) plane,
determined by the astronomical position of the sun (a2-b2).sup.2.
The hypotenuse of triangle ABC can then be generated as a straight
line that simulates the shadow cast, in accordance with the
equation AC.sup.2+BC.sup.2=AB.sup.2.
[0094] This shadow-extraction process is repeated at desired
intervals over an annual time period using the necessary
astronomical data to generate multiple arcs coverages that
represent annual shadow extent. The process is then run as a
programmed data model, at step 915, to generate a set of rendered
shading files. The model repeats this process at desired intervals
(e.g., 6 minutes) over an annual time period using the necessary
astronomical data.
[0095] In step 920, all of the files of rendered shading are
agglomerated (or `dissolved`) into a single file; that is, a mosaic
of the files is created. This rendered shading file contains pixels
that represent the division between shaded and non-shaded areas,
and can be stored in either XY (two-dimensional) or XYZ
(three-dimensional) formats. This single file represents the
complete shading effects from all objects on all objects within the
geographic coverage for one year.
[0096] In step 925, all data in this file that does not represent
shaded areas is eliminated. For example, if the user has chosen to
cast shade over a tri-band color photograph of the coverage, the
shaded areas may be represented by grey-scale/black pixels. In this
case, all pixels that represent color integers (1-256) outside this
grey-scale are eliminated. This new file represents only the XY
coordinates (by pixel) that are shaded at any time or at specific
intervals during the year (or other period of time). Alternatively,
if the user is operating without orthophotographic data, the XY/XYZ
coordinates of the shading areas are separated from those outside
the shaded area in the geographic plane.
[0097] In step 930, the extracted rooftop polygons/area boundary
data are added to the programmable workspace, and rooftop data
files and orientation point data are georeferenced in order to
align the data on the XY plane, in step 935. Next, in step 940, the
XY shadow simulation data (output from step 925) is intersected
with the XY coordinates of the rooftop areas to generate a dataset
of intersected rooftop polygons. All points/pixels that exist
outside these coordinate boundaries are deleted.
[0098] In step 945, areas for the intersected rooftop polygons are
calculated and stored as attributes, which contain the usable roof
area after the intersection process. These calculated areas for
each individual rooftop polygon contain the total non-shaded usable
area per single parcel, and are stored in a rooftop polygon area
file.
[0099] In step 950, the rooftop polygon area file is imported into
the geodatabase 104, with all rooftop FIDs joined to parcel IDs and
georeferenced. An output example of tabular data 955 indicating
geographic coverage is shown below:
[0100] 19 Palms Ave. Miami Fla. 33107--U_AREA 38.652288 m.sup.2
[0101] 21 Palms Ave, Miami Fla. 33107--U AREA 47.576952 m.sup.2
[0102] 23 Palms Ave, Miami Fla. 33107--U AREA 12.930998 m.sup.2
[0103] Additionally, the total usable area may be multiplied by
coordinate-specific solar irradiance levels to provide the actual
output of solar potential for a given roof, as opposed to
indicating the roof area in square meters.
Tabular Data
[0104] In FIG. 3, at step 130, all data is imported into the
geodatabase 104 in tabular form, using GIS 110. In an exemplary
embodiment, the data is calculated and categorized into
parcel-specific tabular data 138, as indicated below.
[0105] Columns for data import contain: [0106] (a) Roof space
(total and portioned) [0107] (b) Roof portion orientation [0108]
(c) Roof portion slope [0109] (d) Usable area for solar
optimization (annual shading on rooftops) [0110] (e) Solar
irradiance/insolation levels (adjusted per square meter)
[0111] These levels are calculated by multiplying the number of
panels available to roof space by irradiance values for the local
zone (e.g., 1280w); [0112] (f) Address and parcel value [0113] (g)
Object ID number (for spatial reference in a GIS) [0114] (h) Parcel
APN number (for municipal legal boundary reference) [0115] (i)
Parcel owner [0116] (j) Roof geometry (gabled, complex etc.) [0117]
(k) Roof material (shingles, tiles etc.)--extracted with
aerial/infrared spectrometer [0118] (l) Return on Investment
(ROI)--estimated ROI for the optimal system available to a
homeowner may be calculated by multiplying irradiance levels per
square meter by available usable roof space/solar panel
model/output, cost of panels, etc.
[0119] Other optional demographics may be imported from other
sources (e.g., household income, marital status, etc.).
[0120] Rows for data import are organized by joining spatially
referenced Object ID numbers to Feature ID numbers, and Parcel
(APN) numbers, as indicated previously. Mathematical calculations,
such as extraction of irradiance levels (e) from usable area (d)
can be completed inside or outside the database, using Microsoft
Excel or other utility program.
[0121] Thematic layers may be used as data inputs to simplify the
geometric relationships within the database. These layers are
utilized to extract items (a) through (I) above. These thematic
layers include some or all of the following: [0122] Municipal land
parcel data, including census block group data, parcel
streets/addresses, and parcel frameworks; [0123]
Aerial/orthophotography, including LiDar and infrared LiDar roof
and clutter data [0124] Digital Elevation Model(s) ground features
as 3D XYZ data; [0125] Geospatial climate dynamics (weather
patterns, rainfall, cloud cover, temperature etc) at
preferred/available grid resolution (e.g., 10 km.sup.2, distributed
point-specific weather station data); [0126] Area-specific
irradiance levels; [0127] Infrared spectrum coverage by remote
sensing; [0128] Estimated return on investment--estimated kW/H per
month of solar potential by parcel as adjusted by state and
municipal rebate programs, hardware costs, etc.
[0129] In FIG. 3, at step 132, GIS database 104 receives spatial
and tabular imports with the following characteristics: [0130] Data
must be standardized with respect to inputs and outputs; and [0131]
Data must be georeferenced correctly to adhere to established
Cartesian spheroid projection.
Cartographic Data
[0132] FIG. 10 is a flowchart showing exemplary details of process
132 (in FIG. 3) for generating cartographic data 139. Initially,
all spatial/vector files created by processes 120, 122, 105, 118,
112, 114, 116 and 104 are imported to geodatabase 104. A
differentiation is made between the attributes of extracted data
(tabular) and the spatial files (cartographic) in this step.
[0133] A greatly simplified sample of a processed vector file 1005
is shown in FIG. 10. In order to create this vector file 1005,
aerial/orthophotographic files of same coverage/coordinates are
imported into database 104. These files are georeferenced using the
same cartographic spheroid (e.g., UTM Zone 17, Clarke 1866) as the
original spatial/vector files. The orthophotographic layer is then
clipped to the spatial boundaries of land parcel data and rooftop
polygons. Object identification numbers are generated for each
clipped orthophotograph and joined to appropriate spatial
identification of the vector files to generate processed raster
file 1010.
[0134] Layers 1005 and 1010 are then overlaid and bound
cartographically in the GIS with spatial attributes for use (at
steps 134 and 135 in FIG. 3) in the DBMS environment, to create
overlaid raster files 1015 with spatial attributes. The inclusion
of raster files facilitates the marketing of the present
technology, and may be considered as part of the overall solar
census process. However, a solar census can be achieved
geometrically without the photography. Aerial photographs are not
required in the geometric extraction process, as they are simply
2-D photos which provide visual assistance for a system user or a
buyer of the service, and should be understood as functionally
separate from the aerial photogrammetry referred to in FIG. 2A.
[0135] In an exemplary ebodiment, raster file data is formatted
into cartographic display 139, and tabular parcel-specific data is
printed out as text 138. Both tabular data and cartographic data
may be filtered at step 134 in accordance with user preferences by
querying GIS database 104.
Generation of Temporal Solar Irradiance Values
[0136] FIG. 11 is a flowchart showing exemplary details of a
process for determining the percentage of total solar irradiance
which is incident on certain areas, at least some of which are
shaded from sunlight during a specific period of time. The present
system follows the process shown in FIG. 11, as described below, to
generate a data set 1135 representing this time-percentage of
shading of a selected area, referred to herein as "solar access"
data. In an exemplary embodiment, data is generated and processed
in raster format. However, the process described herein can be
performed using vector files or other types of files in lieu of
raster files.
[0137] FIG. 12 is a diagram showing exemplary data formats
corresponding to the process shown in FIG. 11. The data initially
used in the present system is taken from Digital Elevation Models
(`DEMs`), and/or equivalent XYZ file formats 142. This data is
collected by means of remote sensing technology. For the purpose of
the present document, it should be noted that a DEM may also be
considered to be a `virtual city` 111, as it is effectively a
three-dimensional rendering of a coverage that includes
three-dimensional man-made features of a city or other area in
which those features are present. However, if the DEM is extracted
for areas that do not contain man-made features, it is not a
requirement for the data to contain such features. DEM data 107
meeting these requirements may be extracted from any digital
elevation model or raw data format representing XYZ coordinates,
regardless of resolution. In an exemplary embodiment, DEM data 107
is in raster (two-dimensional pixel array) format, but DEM data can
also be held in ASCII/text formats, and TINs, etc. Operation of the
present system is best understood by viewing FIGS. 11 and 12 in
conjunction with FIGS. 13-21, as described below.
[0138] FIG. 13 illustrates an example of a Digital Elevation Model
107 extracted from raw XYZ (three-dimensional) data for a specific
area 1400, including a rooftop 1310, where the data is displayed in
a raster format. The Z (elevation) values of the data are contained
within the cells of the DEM, but are only apparent when viewed with
a 3-D display program. As shown in FIGS. 11 and 12, at steps 1101
and 1105 (described in detail below), using DEM raw data 107 (from
step 105, FIG. 2A, or from other sources noted below) `hillshade`
data files 1202 are created for a finite number of points on the
Sun's path (i.e., for the Sun's azimuth and altitude relative to
the location and time points being analyzed) for selected time
intervals or points. The present system extracts individual
temporal solar irradiance values for every point in a particular
coverage area while simulating the Sun's path across the area. Each
hillshade file 1202 is then converted from a signed, continuous
raster surface to a discrete binary format on a cell-by-cell basis.
In an exemplary embodiment, the resultant binary shade file 1202 is
a dataset or file of binary numbers, each representing either a
shadow or a non-shadow condition of a specific area, e.g., a 1 sq.
ft. area at a specific X/Y coordinate, where each file includes
data for a specific point in time at specified XY boundaries. The
data is stored in individual files 1202 as binary code representing
shadow and non-shadow. In an exemplary embodiment, hillshade data
files 1202 are stored as 16-bit integer rasters, although the files
may be stored as ASCII text, other text, or shapefiles, etc.
[0139] Shaded relief, or hill-shading, simulates the cast shadow
thrown upon a raised relief map, or more abstractly, upon the
planetary surface represented. The conventional angle of the light
source for such models is from the northwest (i.e., normally from
the upper left corner of the map). This makes most depictions in
the northern latitudes non-representative of solar light/shadow
patterns, and is done conventionally to avoid multistable
perception illusions (i.e., crater/hill confusion). Because the
viewer is looking at a printed or displayed image, the default
assumption is that light is coming from above. Each hillshade file
1202 is representative of the sun's apparent position on the sky at
a specific time and day of the year.
[0140] As shown in FIG. 11, using astronomical positions of the sun
relative to latitude and longitude (X/Y coordinates) on Earth,
extracted from public records at step 1101, a `hillshade` or shaded
relief file 1202, relative to the irradiance time point, is created
from the raw DEM data 107, at step 1105. At step 1106, hillshade
files 1202 are converted to raw binary data.
[0141] FIG. 14A illustrates an example of a `hillshade` file 1202
for the same area 1400 as shown in FIG. 13, and FIG. 14B shows the
diagram of FIG. 14A, where the layer of data 1202 is converted from
raster format to text format 1401. FIG. 15 shows the superposition
1501 of the raw binary data for hillshade file 1202 over the
original DEM data 107 (of FIG. 13) for area 1400, and FIG. 16 shows
a subset 1601 of the same hillshade data for the area over rooftop
1310.
[0142] Table 1, below, shows an example of the relevant percent of
total daily radiation represented by a given hour for a specific
area, and is generated from the binary values in the hillshade data
in file 1202. The data in Table 1 represents the `curve` of the
Sun's path through the sky. In an exemplary embodiment, the data in
Table 1 is constructed by extracting the files for the latitude and
longitude of a particular area from solar data files prepared by
the National Renewable Energy Laboratory (NREL). These data files
give a historical average of radiation recorded throughout the
country. The data in these files is split in hourly intervals for
every day of the year. By generating a simple distribution curve
from the raw data, the data can be reordered to give a percentage
of total radiation for a given time interval within any other time
interval.
[0143] Each hillshade file 1202 is then `reclassified`, at step
1107, based on irradiance levels representing each selected time
point, e.g., each hour, in the Sun's path for the corresponding
day, to generate `reclassified` shade files 1203. Solar altitude
and azimuth positions for a specific latitude and longitude at a
specific time in a year can be found, for example, at the US Naval
Observatory website (http://aa.usno.navy.mil/data/). If, for
example, azimuth=23.3 and altitude=145.2 on January 15, at
10:00-11:00 PM (shown in Table 2), this time point has a value of
12.67, as indicated in the solar radiation values in Table 1. This
time point value (12.67) represents the value for solar
insolation/radiation during this specific period of time. Every
`reclassified` time point value during a given day thus represents
the percentage of total radiation during that day for a specific
latitude and longitude.
[0144] The conversion chart, or `reclassification` chart created in
step 1107 is shown, in part, in Table 1, below.
TABLE-US-00001 TABLE 1 7:00 8:00 9:00 10:00 11:00 12:00 1:00 2:00
3:00 4:00 January 1.79 7.25 10.50 12.67 13.80 14.05 13.52 12.76
10.99 2.67 February 3.84 7.91 10.09 11.46 12.12 11.38 11.33 9.95
9.95 7.46 March 5.82 8.19 9.83 10.52 11.42 11.46 11.33 10.67 9.33
7.94
[0145] An exemplary set of reclassified files 1203 is shown in
Table 2, below. These files may be considered as layers--there may
be, for example, 10 layers, each of which represents the Sun's
shadow at a certain time for a particular area for every hour from
8 am to 5 pm. Each of these raster layers can be viewed
independently, and each layer has a converted cell value attached
to it for the relevant portion of the day's shade it represents.
Thus, if a shade file is generated for 9 am, and the file
represents 4% of the day's total radiation, every cell which
currently has a value of 1 (where a value of 0=non-shaded, and
1=shaded) is converted in step 1108 (described below) to a `4`.
[0146] The purpose of the reclassified data (in files 1203) is to
provide a conversion of the existing binary data (based on position
and time) to solar radiation (solar access) data. Table 2, below,
shows exemplary reclassified data based on the data in Table 1.
TABLE-US-00002 TABLE 2 Jan. 15.sup.th (San Jose, CA) % Solar
Hillshade File Access Loss Name Being Time AZ AL if Shadded
Reclassified 7:00 -4.6 113.2 1.79 recl_1 8:00 6.1 122.4 7.25 recl_2
9:00 15.5 132.9 10.50 recl_3 10:00 23.3 145.2 12.67 recl_4 11:00
28.9 159.5 13.80 recl_5 12:00 31.5 175.4 14.05 recl_6 13:00 30.7
191.7 13.52 recl_7 14:00 26.8 207.0 12.76 recl_8 15:00 20.2 220.4
10.99 recl_9 16:00 11.6 231.9 2.67 recl_10
[0147] After all of the binary hillshade files 1202 are converted
to reclassified files 1203, the files are then weighted or
normalized and combined by their normalized values, as described
below, to generate a set of normalized (summed/irradiance-weighted)
files 1204, so that the data (which comprises the individual shade
files generated in steps 120-122 of FIG. 3), when summed, has a
total value of 100%.
[0148] FIGS. 17A, 17B, and 17C are exemplary diagrams showing three
sets of binary data 1701/1702/1703 representing hillshade files
1203 for the area over rooftop 1310, after being reclassified from
their counterpart binary hillshade files 1202 to indicate the
respective solar access values. Each of the hillshade files 1203
preferably includes data for one of a selected number of different
time points (e.g., hours) for a selected number of intermediate
intervals (e.g., days) in a selected upper interval (e.g., a
month), although, alternatively, hillshade file data for the same
time points for different intermediate intervals in the upper
interval may be used. The present process comprises the conversion
of a visual file (from a DEM, which has features) to a text/binary
file with no features, only zeros and ones bound to coordinates,
where, for example, a `1` (indicated by the circled value in FIG.
16) is converted to a `7` (indicated by the circled value in FIG.
17A) on one file/layer, and above that layer, another `1` in
another layer at exactly the same point on the grid (representing
area 1400) is converted to, for example, a value of 9 (indicated by
the circled value in FIG. 17B). The same conversion process is
applied to the remaining hillshade files 1202 (a single remaining
file in the present example) in the upper interval to generate file
1703 shown in FIG. 17C.
[0149] Unique values between 1 and 100, representing the percentage
of total radiation during that day for the chosen latitude and
longitude, are calculated in step 1110 (described in detail below).
In effect, a fourth dimension is introduced here. In a single shade
file, the shadow is frozen. However, over a period of time, the
effect of a moving shadow comes into play. When the `weighted`
percent shade value has a zero value, there will always be a zero
in the original binary data in the corresponding `still frame`
hillshade file 1202, because shade never touches this cell,
regardless of the timeline. Thus, a zero in a cell in a one-day
weighted file (e.g., June 12) will mandate a zero in a hillshade at
10 am, but not past the day of June 12th. A zero in a weighted file
for the entire month of June will guarantee a zero on every
hillshade file generated within that month, but not in files in
other months, and so on.
[0150] In step 1108, the set of normalized, reclassified files 1204
is generated by summing groups of reclassified hillshade files
1203, wherein each file 1204 represents the sum of all of the
appropriate normalized files 1203 for a chosen lower interval
within a selected intermediate interval. In the present example,
the lower interval is one hour, and the intermediate interval is
one day, thus each file 1204 represents the sum of all of the files
1203 for one day.
[0151] FIG. 18 is an exemplary diagram showing data 1801 summed
from files 1701, 1702, and 1703 (each of which is a binary
hillshade file 1203) to create normalized file 1204, generated in
step 1108. In an exemplary embodiment, information is extracted for
the percentage of total radiation (i.e., solar access) in one hour
time units which constitute a `lower interval` for the
extraction/processing of data in the present system. This lower
interval is determined by the interval at which hillshade file data
1202 was extracted in step 1101. Hillshade files 1202 created in
step 1105 of FIG. 11 are thus based on one hour intervals in the
present example, although any other time interval may be used.
[0152] The reclassified data 1203 shown in FIGS. 17A-17C now
includes values that can be `summed` or `weighted` in any specified
time interval. Once the desired time intervals are chosen, the
values in the raster files are summed or weighted on a cell-by-cell
(e.g., on a pixel-by-pixel) basis, in step 1110, as described
below. For the purpose of the present description, a `cell` may be
considered to be a single pixel of data. For example, to calculate
the total effect of irradiance for one day, all of the reclassified
values generated within that day are weighted, and the result is a
single raster/file that shows irradiance values for that day. The
same can be done for one month, one season, or one year, etc. These
irradiance-weighted, or normalized raster files 1204 represent the
total (weighted) irradiance for the selected period of time.
[0153] In an exemplary embodiment, the normalized/weighted data is
sorted by using different 60 minute points for each of a selected
number of days of a month. For example, on January 1, data for 8
AM, 9:30 AM, 11 AM, and so on, is generated. On January 15, data
for 8:30 AM, 10 AM, 11:30 AM, etc., is generated, and this process
is repeated for each of the remaining days selected in the month.
This procedure reduces data calculation, and thus requires less
processing. This process can be made more finely-grained to achieve
higher resolution results.
[0154] In an alternative embodiment, direct summing of irradiance
values for every cell is employed, as described above, eliminating
the step of calculating the denominator in equation 1130 (described
below). This process requires only the determination of data
comprising the numerator of equation 1130, based on irradiance
data. This alternative method can be used, for example, for
predicting the total output of a particular solar energy system
over a period of time, and determining related information, such as
ROI (return on investment) over the lifespan of solar cells, etc.
The purpose of this embodiment is to calculate time-dependent
irradiance values which are summed over an interval, without being
averaged.
[0155] In a second alternative embodiment, slope and aspect data is
added on a cell by cell basis to generate irradiance maps which may
include man-made features. For this process, standardized
multipliers are applied over the system outputs. More specifically,
the extracted irradiance value (over time) of each cell is a
function of (and is computed from) the slope value of the same
cell, and the aspect value of the same cell. This process, which
provides increased accuracy for the values per cell over a
particular coverage area, may be implemented using either the
method described immediately above, or using the general method
described in the remainder of this document.
[0156] In step 1110, data for a lower interval 1903 in reclassified
files 1204 is summed for a each of a plurality of intermediate
intervals 1906 within an upper time interval 1910, on a
cell-by-cell basis, to generate a single normalized
(irradiance-weighted) shade raster file 1206 for an upper time
interval 1910 (these intervals are shown in FIG. 19A, described
below). In the present example, the intermediate interval is one
day and there are, e.g., seven intermediate intervals in the upper
interval of one month. The irradiance-weighted raster file 1206
forms the numerator of equation 1130 (shown below), and represents
the upper interval 1910. This intermediate interval is expressed in
units to which the percent values in FIG. 11, step 1101 are
relative. Note that the entire numerator generation process uses a
framework of intermediate interval units. Since the data in
normalized files 1206 is weighted for multiple days, the division
operation (performed in step 1131, as described below) indicated in
equation 1130 is required to adjust the average to correspond to
the selected intermediate interval.
Irradiance Time / repetition interval = Total solar access for
interval Equation 1130 ##EQU00001##
[0157] In the present example, data from sets 1701, 1702, and 1703
is summed in step 1110 to create the output file 1801 (an example
of file 1206) shown in FIG. 18. In the resultant summed/normalized
file 1801, a new value of 25 (indicated by the circled value in
FIG. 18A) is generated for the point on the grid corresponding to
the circled cells in FIGS. 17A, 17B, and 17C. The remaining values
shown in FIG. 18A are generated in like fashion by summing the
respective values in FIGS. 17A-17C.
[0158] The same intermediate time interval is employed consistently
in both the numerator and the denominator of equation 1130. When
creating the numerator in equation 1130, all percent values for
various hours are summed over one or more intermediate time units
(e.g., days) from data sets 1701-1703 within the upper interval
(e.g., a month). Thus, the largest time unit (the upper interval)
for the summing process in step 1110 is the same time unit as the
unit applied for this weighting/summing step of the denominator of
equation 1130.
[0159] Steps 1120-1125 in FIG. 11 show an exemplary process similar
to that described above in steps 1101-1110. In these steps, the
data is assembled based on a hierarchy of specific time intervals.
FIG. 19A is an exemplary diagram showing the generation of a
summed, time-ranked hillshade frequency file 1207. FIG. 19B is an
exemplary data model showing the process of composite-to-new raster
generation, further detailing steps 1120 and 1125 in FIG. 11.
Operation of the present system is best understood by viewing FIGS.
19A and 19B in conjunction with one another.
[0160] As shown in FIG. 19A, lower intervals 1903 of data, i.e.,
data representing time units smaller than a corresponding
intermediate interval 1906 (a week in the present example), may be
considered as being contained inside each intermediate interval
1906. Data is grouped in intermediate intervals within a larger
upper interval 1910, e.g., a month. In FIG. 19A, lower time
interval 1903 is shown as being equivalent to the interval
represented by each of the hillshade files 1202'. These lower
intervals 1903, by which hillshade files 1202' are grouped within
each file 1205, can be minutes, hours, or any other units of time
which are smaller than the intermediate time interval. In the
present example, these lower intervals 1903 are hours, and the
intermediate interval 1906 represents a `container` in which data
for lower intervals of hillshade files 1202' are collected for
layering. If the intermediate interval is a day, there are up to 24
(or 14 for daylight hours) hourly hillshade files that are layered.
If the upper interval 1910 is a month and there is one hillshade
file 1202' generated for noon every day, then 30 hillshade files
1202', in binary format, are layered and summed. The composite
intermediate interval files 1205 are grouped as indicated by the
sets of files 1205 in grouping 1910 shown in FIG. 19A.
[0161] In the present example, as shown in FIG. 19B, the frequency
(per lower interval 1903) of hillshade files 1202/1202' is daily.
Therefore, as shown in FIG. 19B, a frequency value 1904 of `hourly`
is input to generator 1905 at step 1901. The intermediate interval
1906 in the time interval hierarchy is then chosen, at step 1902.
In the present example, the intermediate interval is one day.
[0162] As shown in FIGS. 11, 19A, and 19B, once the preferred
intermediate interval 1906 is determined, all hourly, or other
lower interval, hillshade files 1202' generated in steps 1105/1106
for the lower time interval are layered (i.e., corresponding cell
values in each file are inclusively `OR`-ed into a single composite
file 1205 by generator 1909, in step 1120. In step 1120, these
files are grouped via whatever sorting process is desired. For
example, if shade files have been generated for every hour of the
day for 7 days in the month, and it is desired to extract a
percentage of total daily radiation lost from shading effects for
that month, the files are grouped into sets representing single
days. Thus, in the example shown in Tables 1 and 2, single hourly
interval data for January 15 is grouped into each of 7 files, and
at step 1120 a `composite` is formed from the files to create
composite hillshade frequency file 1207 having decimal (i.e.,
non-binary) data for each day in the chosen upper interval. This
new file 1207 represents all spatial (X,Y) coverage for shaded and
non-shaded areas in each of the intermediate intervals (e.g., each
day of seven selected days in the month). Note that the initial
data (set 1201) used in step 1120 is the same as the data 122 from
step 120 in FIG. 2A.
[0163] At this point, an upper interval set 1910 of binary files
1205 (where, e.g., 0=non-shaded, 1=shaded) has been created for
each intermediate interval 1906 (e.g., a day), where each file 1205
is a composite (a layered data set) of hillshade files 1202'
representing `L` lower intervals (e.g., hours) constrained by the
intermediate interval size 1906 (one day in the present case) and
representing the total XY extent of solar access/shade effects over
the time period chosen.
[0164] In the present example, the goal is to determine the average
daily solar irradiance values, or solar access, for a particular
month for a given set of cells. Now that data for each of the days
(the intermediate interval) have been layered into a set of files
1910, the daily values, contained within a single file 1205
representing binary shade values for the entire day, are weighted
on a cell by cell basis to create the summed frequency file 1207
containing weighted values within the upper interval, which is one
month in the present example.
[0165] Once a composite file 1205 (which remains as per-cell binary
code) for each of the days in intermediate interval 1906 is
created, these intermediate interval files are summed, at step
1125. The result is a single summed shade/time frequency file 1207
in which each cell represents the frequency of repetition of small
interval files for the chosen upper time interval. Cells range in
value from 0 to N, where N is the number of days over which an area
is being evaluated.
[0166] Summed shade/time frequency file 1207 (which is no longer a
file with binary data, but one containing decimal [base 10] integer
values) generated in step 1125 provides the frequency of hillshade
file overlap or `frequency of repetition` for every cell in the
coverage that was summed in the numerator of equation 1130. With
the creation of the denominator of equation 1130, i.e., file 1207,
cells in the denominator file (in the present example) will have
one of four possible values--a value of 0, indicating that the area
represented by the cell is never shaded, or values of 1, 2, or 3,
indicating the frequency of repetition of corresponding hillshade
small interval time point data within the selected upper
interval.
[0167] Therefore, at step 1131, the output from step 1110 (file
1206) is divided by the output of step 1125 (file 1207), using a
matrix division operation, to provide the average percent of solar
access per cell for the period of the upper interval. The outputs
from steps 1110 and 1125 are operated on in accordance with
equation 1130:
Irradiance ( not time adjusted ) Repetition interval = Total solar
access for interval ##EQU00002##
[0168] The output of equation 1130 represents the solar access, or
total percentage of daily radiation shade (per cell) 1135 for the
upper time interval 1910 over the entire coverage area that has
been analyzed.
[0169] FIG. 20 is an exemplary diagram showing summed shade/time
frequency file 2001 (an example of file 1207) generated from file
sets 1701, 1702, and 1703, representing the denominator of equation
1130. FIG. 20 represents the denominator of equation 1130. In step
1131, each cell in file 1206 (of which file 1801 in FIG. 18 is an
example) is divided by the corresponding values in file 1207.
[0170] In the present example, the frequency of repetition of
corresponding hillshade small interval time point data within the
selected upper interval is either once in a month, or three times a
month, as indicated by the "1"s and "3"s in FIG. 20, which
correspond to the sum of the number of non-zero-value cells in all
of the hillshade files 1203.
[0171] FIG. 21 is an exemplary diagram showing resultant data 2101
after division operation 1131 has been performed to generate
corresponding file 1135, as described above. In the present
example, each of the cells in file 1206 has been divided by the
(value of the) corresponding cell in file 2001/1207, which
represents the frequency of repetition (overlap) of small interval
files 1203 for the selected upper time interval. The circled cell
in file 1801 (in FIG. 18) having a value of 25 is thus divided by a
value of 3, (as shown in FIG. 20) to determine the monthly average
for the cell, which is 8.3, as indicated in file 2101 in FIG.
21.
[0172] Certain changes may be made in the above methods and systems
without departing from the scope of that which is described herein.
It should be noted that the present method is not limited to the
extraction of data on rooftops. All matter contained in the above
description and shown in the accompanying drawings is to be
interpreted as illustrative and not in a limiting sense. For
example, the methods shown in the accompanying drawings may include
steps other than those shown therein, and the systems shown in the
drawings may include different components than those shown therein.
The elements and steps shown in the present drawings may be
modified in accordance with the methods described herein, and the
steps shown therein may be sequenced in other configurations
without departing from the spirit of the system thus described. The
following claims are intended to cover all generic and specific
features described herein, as well as all statements of the scope
of the present method, system and structure, which, as a matter of
language, might be said to fall there between.
* * * * *
References